Evaluation of RANS and LES turbulence models for indoor airflows

In buildings low indoor airflows persist, which presents difficulties in modelling of the airflow due to its transitional flow behaviour. The application of a number of turbulence models are investigated in order to evaluate their performance and to provide some issues related to indoor room modelling. The most basic turbulence models are the RANS based models which are quick in providing a solution but are known to fail in transitional flows. Under different conditions such as a channel flow (to predict near wall turbulence), and free jets (to predict diffuser ventilation dispersion), the standard k-e (SKE) model surprisingly provided sufficient results. Further experimental data was produced which allowed a calibration of the SKE model to further improve its perfonnance. Because the SKE model is so widely used and is known for its generality, the calibrated model was further tested using an enclosed room, and was compared with a Large Eddy Simulation model and experimental data. Finally, the instantaneous results from the LES model highlighted the deficiency of steady simulations using RANS turbulent models which average and smear out the effects of local recirculating eddies.